Overgrowth of a leukemic culture by a minor CD34+ population

Leukemia Research 22 (1998) 549–556

Overgrowth of a leukemic culture by a minor CD34+ population

Paul Baines *, Helen Lake, Janet Fisher, Louise Truran, Terry Hoy, Alan K. BurnettHaematology Department, Uni6ersity Hospital of Wales and Uni6ersity of Wales College of Medicine, Cardiff CF4 4XW, UK

Received 23 September 1997; accepted 26 January 1998

Abstract

We have investigated the differentiation potential of blast cells in a case of acute myeloid leukemia which comprised a majorityCD34− population and a minor (2%) CD34+ fraction. Blasts were cultured for 2 weeks in a combination of cytokines—c-Kitligand, interleukin 3 and granulocyte–macrophage colony-stimulating factor (SIGm mix)—together with all-trans retinoic acid or1a,25-dihydroxy vitamin D3. Maturation of blasts was assessed by morphology on Romanowsky-stained slides, changes in surfaceCD markers and clonogenic culture. After 7 days of culture of unseparated blasts in SIGm, most maturation was monocytic, butwith retinoic acid 63% of blasts had matured into granulocytes. Vitamin D3 enhanced monocytic differentiation, with 60% of cellsbecoming monocytic. The percentage of CD14 and CD15 positive cells decreased over 7 days in SIGm (from 62% to 17% andfrom 76% to 39% for CD14 and CD15, respectively). CD14+ cell numbers were maintained, or recovered, in culturessupplemented with vitamin D3 (59% at day 7), and CD15+ cell numbers, too, remained unchanged in the presence of retinoicacid (67%) or vitamin D3 (66%). Aberrant markers CD7 and CD56 declined under any conditions. When separated, both theCD34− and CD34+ fractions showed similar changes in morphology and surface maturation markers, suggesting that these twopopulations may be closely related. However, only a few CD34+ cells expressed the aberrant markers present on the majorityblast population. The CD34− population declined in culture while the CD34+ fraction rapidly expanded. This probably reflectsthe difference in progenitor content; high numbers of colony-forming cells were concentrated in the CD34+ subpopulation. Weconclude that both CD34− and CD34+ populations can differentiate but only the CD34+ fraction proliferates. Primitiveclonogenic CD34+ cells from this patient may generate occasional aberrant CD34+ blasts which could then differentiate intothe accumulating aberrant CD34− blast population. © 1998 Elsevier Science Ltd. All rights reserved.

Keywords: AML; Cytokines; CD markers; CFC; Differentiation; Inducers

1. Introduction

Differentiation therapy [1–3] has attracted attentionas an alternative or an adjunct to cytotoxic chemother-apy for acute myeloid leukemia (AML). Amongst thelikely differentiation agents are the retinoids and vita-min D3 (vitD3), both of which have proved promisingcandidates for differentiation therapy from in vitrostudies on cell lines [4,5]. However, apart from acutepromyelocytic leukemia, which responds well to all-trans retinoic acid (ATRA) [6–8], other AML subtypesare only occasionally responsive to these inducers [9–11].

We report here a leukemia which differentiates wellwhen ATRA and vitD3 are added to cytokine-supple-

mented cultures of AML blasts. We routinely add amixture (SIGm mix) of stem cell factor, interleukin 3and granulocyte–macrophage colony-stimulating factor(GM-CSF) to our cultures. These are cytokines shownby others to support the proliferation of AML blasts[12–14], though not their differentiation [10,15].

An unusual feature of this AML was the high yieldof colony-forming cells which belonged to a minorityCD34+ fraction with high proliferation capacity.There is some evidence [16] that CD34+ subpopula-tions in AML bear aberrant cytogenetic markers sug-gesting that minor stem cell fractions can be leukemic.Indeed, more recent data show that CD34+ +CD38− fractions from several AMLs were clearlycapable of transplanting leukemia to SCID and NOD/SCID mice [17].

Observations like these led us to compare the in vitrobehaviour of the CD34+ and CD34− populations in

* Corresponding author. Tel.: +44 1222 743486; fax: +44 1222744655.

0145-2126/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved.

PII: S0145-2126(98)00043-5

P. Baines et al. / Leukemia Research 22 (1998) 549–556550

order to evaluate their likely contribution to malig-nancy in our leukemia.

2. Materials and methods

2.1. Clinical details of the patient

The patient presented with chronic myelomonocyticleukemia at 77 years of age and a few months laterdeveloped AML M4/M5 as determined by marrowmorphology and CD marker analysis of blastpopulations.

2.2. Patient samples

Patient’s blood was collected into preservative-freeheparin (10 units ml−1) and mononuclear cells (MNC)prepared by centrifugation over lymphocyte separationmedium (Gibco). MNC were washed once in Hepes-buffered minimal essential medium (MEM; Gibco) andresuspended in MEM supplemented with 40% foetalbovine serum (Gibco or Imperial) and 10% dimethyl-sulphoxide (Sigma) before being placed in a Cryo 1freezing container (Nalgene) at room temperature andallowed to cool to −70°C. Cryovials were thawedquickly and diluted dropwise, slowly, using MEM con-taining 10 units ml−1 preservative-free heparin, andresuspended in MEM for dilution into culture. MNCcontained less than 15% non-blast cells. Cells wereseeded at 106 ml−1 in liquid culture and between 5×104 and 5×105 ml−1 for semi-solid culture.

2.3. CD34 selections

Blasts were resuspended in MEM+1% bovine serumalbumin (BSA) at 2.5–6×106 cells ml−1 culture.MEM–BSA was used as a washing medium through-out. MNC were mixed with CD34-coated Dynabeads(coated with mAb 561, recognizing a class III epitope;Dynal, UK) at 4×107 beads ml−1 cell suspension.After rotation at 4°C for 30 min, rosetted CD34+ cellswere separated using a Dynal MPC magnet; cells still insuspension were decanted as the CD34− fraction.Detachabead CD34 (Dynal, UK) was added to theCD34+ cells for 15 min at 37°C; the beads were thenmagnetically removed from the cell suspension.CD34+ and CD34− cells were added to liquid andsemi-solid cultures at 103 cells ml−1 and 5×105 ml−1,respectively.

2.4. Liquid and semi-solid cultures

Iscove’s Modified Dulbecco’s Medium (Gibco) wasmade to 600 mOsm and supplemented with antibiotic–antimycotic (15240-039; Gibco), 32 ml/ml 7.5% sodium

bicarbonate (Gibco), 2% deionised BSA (fraction V;Sigma), 5×10−5 M 2-mercaptoethanol (BDH), 360 mgml−1 human transferrin (Boehringer Mannheim) and20% foetal calf serum (Imperial). This mix was dilutedwith a one-third volume of double-distilled water or 1%Difco Noble agar (in water) for liquid and semi-solidculture, respectively. Liquid cultures were performed in25 ml flasks (Bibby Sterilin), and semi-solid cultures in24 multiwell plates (Nunc). Liquid cultures were re-fedby demipopulation after 7 days. Semi-solid cultureswere scored on a dissecting microscope after 2 weeks,with white colonies (more than 50 cells) being scored asgranulocyte–monocyte and red colonies as erythroid.Clusters are clones of 5–50 cells. Colonies were pluckedfrom 2-week semi-solid cultures using a narrowed glasspasteur pipette and dispersed into 100 ml of MEM forFACS analysis.

2.5. Purified, recombinant cytokines

Stem cell factor (Amgen, Thousand Oaks, CA) ormast cell growth factor (Immunex, Seattle, WA) wereused as sources of c-Kit ligand, both at 20 ng ml−1

final concentration. Interleukin 3 was kindly donatedby Genetics Institute (Cambridge, MA) and used at 5ng ml−1. GM–CSF was a gift from Immunex and wasused at 5 ng ml−1. Recombinant erythropoietin(Boehringer Mannheim, Germany) was used at 2 unitsml−1 of culture.

2.6. Inducers

All-trans retinoic acid (Sigma) and 1a,25-dihydroxyvitamin D3 (Calbiochem) were both dissolved in abso-lute ethanol as a 10−4 M stock. An appropriate dilu-tion of ethanol was added to control cultures.

2.7. Morphology

Slides of cells from liquid culture were prepared on aShandon Cytospin 2, and these were stained usingconventional Romanowsky dyes.

2.8. Surface markers

The following fluorescence-conjugated, monoclonalantibodies from commercial sources were used as sin-gle, double or triple surface markers: CD7FITC(Dako), CD7PE (Caltag), CD14FITC (Dako),CD14PE (Becton Dickinson) CD15FITC (Becton Dick-inson), CD56PE (Becton Dockinson), CD34Cy5(Dako), Simulset Leucogate (CD45FITC, CD14PE;Becton Dickinson). Negative controls includedIgG1FITC (Dako), IgM (Dako), IgG1PE (BectonDickinson) and IgG1CY5 (Dako). Cells from liquidcultures, or plucked colonies, were suspended in 100 ml

P. Baines et al. / Leukemia Research 22 (1998) 549–556 551

Fig. 1. Surface marker expression on differentiating leukemic blasts after 7 days of culture in cytokines without (SIGm) or with retinoic acid(SIGmRA) or vitamin D3 (SIGmD3).

of MEM at 107 ml−1 from liquids. The 100 ml of cellswere washed with PBS–BSA, resuspended in residualbuffer and stained with one, two or three antibodies for30 min at 4°C. Cells were then washed twice withPBS–BSA and resuspended in 300 ml of 2%paraformaldehyde before acquisition on a Becton Dick-inson FACScan (within 2 days). PC Lysys and WINMDI software were used to analyse data, setting for-ward and perpendicular light scatter gates around vi-able cells.

2.9. Statistical tests

Non-parametric (Mann–Whitney) or paired t-testswere used.

3. Results

The following data on liquid cultures were selectedfollowing repeat thawings (2–5) of frozen blasts toensure consistent results.

3.1. Cellularity

In the presence of the SIGm cytokine combination,cell numbers were maintained (Fig. 1) over the firstweek (81% of the cells introduced into culture) andnumbers increased 15-fold over the ensuing 7 days. Onaddition of retinoic acid, cellularity was 79% at day 7with an 8-fold increase in the second week. For cultureswith vitD3 the respective values were 88% and 10-fold.

3.2. Surface markers

The proportion of CD14+ and CD15+ cells de-creased in SIGm from 62% to 17% and from 76% to39%, respectively, over the first week (Fig. 1). A similarfall in CD14+ cells occurred with ATRA but CD15+cell numbers were significantly higher (67%; p=0.021;n=4) when this inducer was added to SIGm cultures.CD15+ cell numbers were also maintained in thepresence of vitD3 (66%), as were CD14+ cells (59%).

Both CD7 and CD56 were lost from unseparatedblasts over 1 week of SIGm culture, declining from 80%to 23% and from 73% to 16%, respectively (Fig. 1). Thisloss also occurred in the presence of inducers.

3.3. Morphology

Over 7 days of SIGm culture of unseparated blasts,nucleated cell numbers fell to 33%, with a rise inmonocyte numbers from 28% to 43%. Maturing granu-locytes accounted for only 6% of all cells. ATRAsignificantly (p=0.03, n=4) increased the proportionof granulocytes to 63% (14% myelocytes, 49%metamyelocytes/neutrophils). With vitD3, 60% of cellsmatured to monocytes.

3.4. Clonogenicity

Scores of 48000 granulocyte-monocyte colony form-ing cells (GM-CFC) and 20000 erythroid colony form-ing cells (E-CFC) ml−1 blood were obtained from thispatient. In culture, GM-CFC numbers increased 27-,34- and 60-fold, respectively, over 7 days in SIGm,


Fig. 2. Control (a, c) and test (b, d) plots of unseparated leukemic blasts labelled with antibodies to (b) CD34(CY5)/CD7(FITC-horizontal) and(d) CD34(CY5)/CD56(PE-horizontal), showing that the bulk of the minority CD34+ population (arrowed) lacks CD7 and CD56.

SIGm+ATRA or SIGm+vitD3. Erythroid coloniesdisappeared.

3.5. Origin of progenitors

No cytogenetic or molecular abnormality was presentin this patient’s cells that could be used to determinewhether the clonogenic cells represented an expansion ofnormal or preleukemic progenitors. When gated from

triple-labelled unseparated blasts on the FACS, mostCY5-labelled CD34+ cells did not express the aberrantmarkers (CD7 and CD56) characteristic of the bulk blastpopulation (Fig. 2) although a few did (Fig. 3). Twelvecolonies plucked from semi-solid cultures of blasts failedto yield cells with aberrant CD7 or CD56 (Fig. 4). Notethat the apparent presence of CD56 on some CD15+cells is a distortion arising from inadequate compensa-tion for the very bright CD15 expression.


CD45/CD14 Leucogate demonstrated the presence ofboth granulocytes and monocytes in 12 individualplucked myeloid colonies (Fig. 5) but also granulocytesin 3/3 plucked erythroid colonies (Fig. 6). Only 26% ofcells in plucked erythroid colonies stained forglycophorin.

3.6. CD34 selection

The following figures are the mean of two CD34separations. Only 0.2% of all blasts added toDynabeads were recovered in the CD34+ population.

Fig. 4. Lack of aberrant CD56 on plucked colonies from SIgmD3cultures of blasts from patient. Top plate shows CD15 versus CD56staining on cells present in a block of background agar and accumu-lated over the same time as the two lower plates.

Fig. 3. Mononuclear cells triple-stained for CD34(CY5) and aberrantmarkers CD7 and CD56. Plots of CD7(FITC-horizontal) versusCD56(PE) on (a) gated CD34− blasts and (b) gated CD34+ cellsshowing that a small proportion of these cells (arrow) express aber-rant markers.


Fig. 5. Leucogate (CD14 versus CD45) staining on plucked myeloidcolony demonstrating the presence of monocytes (m) and granulo-cytes (g).

3.7. Clonogenicity of blast fractions

Few colony-forming cells were recovered in CD34−fractions selected from uncultured blasts; GM-CFC andE-CFC amounted to only 0.3% and 0.6% of cells in theCD34− fraction. In contrast, GM-CFC and E-CFCcomprised 4.7% and 4.3% of cells in the minorityCD34+ fraction.

3.8. Liquid culture of CD34-separated fractions

In liquid SIGm culture of CD34− cells from the twoCD34 experiments, only 16% and 4.8% (mean 10.4%)of nucleated cells seeded, and no colony-forming cellswere recovered after 1 week. However, nucleated cellsexpanded 20-fold and 2-fold (mean 11-fold) from thetwo prepared CD34+ fractions; SIGm+ATRA andSIGm+vitD3 induced 43-fold and 11-fold mean in-creases, respectively. Under the three conditions, GM-CFC numbers expanded 18-, 19- and 66-fold,respectively, from the CD34+ fraction.

Taking the first of the CD34 fractionation experi-ments, CD34− cells declined to 16% in 7 days, whileCD34+ cells proliferated 20-fold. For every 105 cellsseeded, only 1.6×104 CD34− cells would be left after1 week while CD34+ cells would expand 20-fold, from0.2×104 to 4×104 and would now form 70% of allcells present. These figures show that the decliningCD34− population would soon be overgrown by pro-liferating CD34+ cells. The values calculated from thesecond CD34 experiment indicate that 5% of CD34−cells survived to day 7, with CD34+ cells increasing to44% of the total. After 2 weeks of SIGm culture,CD34− cells would have declined to 12% of the origi-nal number seeded while CD34+ derived cells wouldhave increased 60-fold to contribute 90% of all cellspresent.

Both fractions showed similar changes in morphol-ogy to those seen in cultures of unseparated blasts. Forexample, ATRA increased maturing granulocytes inSIGm cultures of CD34+ cells from 0% to 40%.Granulocytes increased with ATRA from 24% to 60%in SIGm cultures of CD34− cells. CD changes weresimilar too, as illustrated in Table 1.

4. Discussion

The pronounced response to differentiation inducers,the massive cellular expansion and high progenitorcontent marked this leukemia out from the majority ofour AML samples.

For therapy, the crucial question is whether theCD34+ cells and progenitors are part of the frankleukemia or represent a preleukemic subpopulation orare some kind of reactive polyclonal haemopoiesis.

This is 10-fold fewer than the number of CD34+ cells(2%) estimated by FACS analysis of leukemic blast cellpreparations and represents a 10% recovery of availableCD34+ cells. Since few clonogenic cells were recoveredin the CD34− fraction it seems likely that manyCD34+ cells remain attached to the beads.

Fig. 6. Leucogate staining on plucked erythroid colony showing thepresence of granulocytes (g).


Table 1CD surface marker changes in CD34 separated fractions cultured inSIGm with or without inducers

Fraction InducerDay % Cells positive for:

CD14 CD15

CD34− 0 – 54 707 – 10 63

ATRA 97 797 VitD3 92 79

– 150 NDCD34+7 – 22 367 ATRA 0 91

VitD3 82 927

ATRA, all-trans retinoic acid; VitD3, vitamin D3.ND, not done for lack of cells.

blasts lose their aberrant markers anyway. The expres-sion of aberrant markers on leukemic cells is fairlyephemeral with both loss and gain being reported in theliterature [28–30].

Probably because so many progenitors were present,CD34+ cells quickly outgrew CD34− blasts in ourculture system. This resembles the reported regenera-tion of the haemopoietic system by minority leukemicCD34+ cells in NOD/SCID and SCID mice [17,31]although these mice ultimately succumb to leukemia.We have no evidence of re-emergence of leukemia inour cultures but perhaps more time is required for thisto occur or the differentiation pressure may be toostrong in vitro.

In conclusion, the CD34− and CD34+ fractionsshare closely related maturation responses to inducersbut their precise relationship remains unproven. Bear-ing in mind the recent literature [16,17,27], the CD34+progenitor population may contain leukemic stem cells.Could these be the CD34+ cells which express aber-rant markers or are such cells merely destined to con-tribute to a non-proliferating and accumulatingCD34− blast pool? One way of resolving this conun-drum might be to search for aberrant cells in minorityCD34+ populations in AML to see if they are theones with leukemic potential in NOD/SCID mice.Whatever the link, while the presence of large numbersof abberant-free progenitors might assist regenerationfollowing chemotherapy or transplantation, there is therisk that such progenitors could occasionally generateleukemic cells, which would eventually lead to relapse.

Acknowledgements

We are indebted to the clinical staff for providingblood samples, to Steven Couzens (Immunophenotyp-ing Section) for his prompt replies to our enquiries intoCD marker expression on patients’ blasts, to A. Agerwho prepared several of the frozen samples and to DrSeah Lim for his helpful comments.

References

[1] Sachs L. Control of normal cell differentiation and the pheno-typic reversion of malignancy in myeloid leukaemia. Nature1978;274:535.

[2] Koeffler HP. Induction of differentiation of human acute myel-ogenous leukemia cells: therapeutic implications. Blood1983;62:709.

[3] Olson I, Bergh G, Ehincer M, Gullberg U. Cell differentiation inacute myeloid leukemia. Eur J Haematol 1996;57:1.

[4] Breitman TR, Selonick SE, Collins SJ. Induction of differentia-tion of the human promyelocytic leukemia cell line (HL60) byretinoic acid. Proc Natl Acad Sci USA 1980;77:2936.

[5] Brackman D, Lund-Johansen F, Aarskog D. Expression of cellsurface antigens during the differentiation of HL-60 cells induced

Residual normal progenitors have occasionally beendescribed in AML [18,19], and we have already demon-strated that the culture conditions used here will sup-port the proliferation of normal CD34+ cells withextensive granulocytic maturation in the presence ofATRA [20]. However, the history of myelodysplasia inthis patient, together with the elevated (often multipo-tent) progenitor levels which are a feature of some casesof chronic myelomonocytic leukemia (CMML) [21],makes it more likely that the progenitors, and theCD34+ fraction to which they belong, arepreleukemic.

If so, this raises the question of whether this minoritypopulation currently feeds into the blastic phase of theleukemia or whether a self-renewing blast stem cellpopulation has evolved from the CMML clone, andnow dominates in vivo. Cytogenetic abnormalities havebeen reported in minor CD34+ fractions in leukemia[16] and clonal genetic markers can be detected inerythroid and lymphoid fractions from AML andmyelodysplastic patients [22–26]. These observationssuggest that primitive cells can be part of the leukemicprocess [27]. This is apparently confirmed by recentreports that minority CD34+ + CD38− cells fromAMLs could transplant leukemia in NOD/SCID mice[17].

No cytogenetic marker was present to clarify the roleof CD34+ cells in our leukemia. CD34− blasts showsimilar morphological changes to the CD34+ derivedcells in response to inducers, and CD14 and CD15acquisition by both fractions was enhanced by ATRAand vitD3. These similarities suggest that the two popu-lations are closely related.

Very few CD34+ cells expressed aberrant markersand these could conceivably be about to mature intoaberrant CD34− blasts. The rarity of such cells mightexplain why we failed to detect them in pluckedcolonies but it seems, from our data, that cultured


by 1,25dihydroxyvitamin D3, retinoic acid and DMSO. LeukRes 1995;19:57.

[6] Chomienne C, Ballerini P, Balitrand N, Daniel MT, Fenaux P,Castaigne S, Degos L. All-trans retinoic acid in acute promyelo-cytic leukemia. II. In vitro studies: structure–function relation-ship. Blood 1990;76:1710.

[7] Menger H, Yu-chen Y, Shu-rong C, Jin-ren C, Jia-Xiang L, LinZ, Long-jun G, Zhen-yi W, Wang Z-Y. Use of all-trans retinoicacid in the treatment of acute promyelocytic leukaemia. Blood1988;72:567.

[8] Chen ZX, Xue YQ, Zhang R, Tao RF, Xia XM, Li C, Wang W,Zu WY, Yao XZ, Ling BJ. A clinical and experimental study onall-trans retinoic acid-treated acute promyelocytic leukemiapatients. Blood 1991;78:1413.

[9] Ossenkoppele GJ, Wijermans PW, Nauta JJP, Huigens PC,Langenhuisjen MMAC. Maturation induction in freshly isolatedhuman myeloid leukemic cells, 1.25(OH)2 vitamin D3 being themost potent inducer. Leuk Res 1989;13:609.

[10] Kelsey SM, Makin HLJ, Newland AC. Functional significanceof induct:ion of differentiation in human myeloid leukaemicblasts by 1,25-dihydroxyvitamin D3 and GMCSF. Leuk Res1992;16:427.

[11] Sakashita A, Kizaki M, Pakkala S, Schiller G, Tsuruoka N,Tomosaki R, Cameron JF, Dawson MI, Koeffler HP. 9-cisretinoic acid: effects on normal and leukemic hematopoiesis invitro. Blood 1993;81:1009.

[12] Goselink HM, Williams DE, Fibbe WE, Wessels HW, Bever-stock GC, Willemze R, Falkenburg JHF. Effect of mast cellgrowth factor (c-kit ligand) on clonogenic leukemic precursorcel]s. Blood 1992;80:750.

[13] Pebusque M-J, Fay C, Lafage M, Sempere C, Saeland S, CauxC, Mannoni P. Recombinant human IL3 and GCSF act syner-gistically in stimulating the growth of acute myeloid leukemiacells. Leukemia 1989;3:200.

[14] Miyauchi J, Kelleher CA, Wong GG, Yang Y-C, Clark SC,Minkin S, Minden MD, McCulloch EA. The effects of combina-tions of the recombinant growth factors GM-CSF, G-CSF, IL3and CSF-1 on leukemic blast cells in suspension culture.Leukemia 1988;2:382.

[15] Salem M, Delwel R, Mahmoud LA, Clark S, Elbasousy EM,Lowenberg B. Maturation of human acute myeloid leukaemia invitro: the response to five recombinant haematopoietic factors ina serum-free system. Br J Haematol 1989;71:363.

[16] Mehrotra B, George TI, Kavanau C, Avet-Loiseau H, Moore D,Willman CL, Slovak ML, Atwater S, Head DR, Pallavicini MG.Cytogenetically aberrant cells in the stem cell compartment(CD34+ lin− ) in acute myeloid leukemia. Blood 1995;86:1139.

[17] Bonnet D, Dick JE. Human acute myeloid leukemia is organizedas a hierarchy that originates from a primitive hematopoieticcell. Nat Med 1997;3:730.

[18] Turhan AG, Lemoine FM, Bonnet ML, Baillou C, Picard F,Macintyre EA, Varet B. Highly purified primitive hematopoleticstem cells are PML–RARA negative and generate nonclonalprogenitors in acute promyelocytic leukemia. Blood1995;85:2154.

[19] Raymakers R, Wittebol S, Pennings A, Linders E, Poddighe P,De Witte T. Residual normal, highly proliferative progenitorscan be isolated from the CD34+/CD33− fraction of AMLwith a more differentiated phenotype (CD33+ ). Leukemia1995;9:450.

[20] Truran L, Baines P, Hoy T, Burnett AK. GCSF augmentspost-progenitor proliferation in serum-free cultures of myelodys-plastic marrow while ATRA enhances maturation. Leuk Res1998;22:241.

[21] Geissler K, Hinterberger P, Bettelheim P, Hass O, Lechner K.Colony growth characteristics in chronic myelomonocyticleukemia. Leuk Res 1988;12:373.

[22] Tefferi A, Thibodeau SN, Solberg LA. Clonal studies in themyelodysplastic syndrome using X-linked restriction fragmentlength polymorphism. Blood 1990;75:9.

[23] Butturini A, Gale R P. Relationship between clonality andtransformation in acute leukemia. Leuk Res 1991;15:1.

[24] Sun G, Wormeley S, Sparkes RS, Naeim F, Gale RP. Wheredoes transformation occur in acute leukemia. Leuk Res1991;15:1183.

[25] Greaves M. Stem cell origins of acute leukemia and curability.Br J Cancer 1993;67:413.

[26] Cuneo A, Balboni M, Carli MG, Bigoni R, Roberti G, Pazzi I,Previati R, Castoldi G. Involvement of erythrocytic and granu-lomonocytic lineages by trisomy 11 in two cases of acutemyelomonocytic leukemia with trilineagemyelodysplasia. CancerGenet Cytogenet 1994;77:33.

[27] Geller RB, Bray RA. CD34 expression on acute myelocyticleukemia cells. commentary. Leuk Res 1997;21:387.

[28] Campos L, Guyotat D, Larese A, Mazet L, Bourget JP, ErshamA, Fiere D. Expression of CDl9 antigen on acute monoblasticleukemia cells at diagnosis and after TPA-induced differentia-tion. Leuk Res 1988;12:369.

[29] Traweek ST, Ben-Ezra J, Braziel RM, Winberg CD. The in-vitroresponse of CD2-positive acute myelogenous leukemia to prolif-eration and differentiation inducing agents. Leuk Res1990;14:433.

[30] Tien H-F, Wang C-H, Chen YC, Shen M-C, Lin D-T, Lin K-H.Characterization of acute myeloid leukemia (AML) coexpressinglymphoid markers: different biologic features between T-cellantigen positive and B-cell antigen positive AML. Leukemia1993;7:688.

[31] Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, Caceres-Cortes J, Minden M, Paterson B, Callgiuri MA, Dicke JA. A cellinitiating human acute myeloid leukemia after transplantationinto SCID mice. Nature 1994;367:645.

.

Overgrowth of a leukemic culture by a minor CD34+ population

Documents

Transcript of Overgrowth of a leukemic culture by a minor CD34+ population